Download Advances in Multimedia Modeling: 16th International by Shi-Kuo Chang (auth.), Susanne Boll, Qi Tian, Lei Zhang, PDF

By Shi-Kuo Chang (auth.), Susanne Boll, Qi Tian, Lei Zhang, Zili Zhang, Yi-Ping Phoebe Chen (eds.)

The sixteenth overseas convention on Multimedia Modeling (MMM2010) was once held within the recognized mountain urban Chongqing, China, January 6–8, 2010, and hosted through Southwest collage. MMM is a number one foreign convention for researchersand practitioners to percentage their new principles, unique study effects and practicaldevelopment reviews from all multimedia comparable parts. MMM2010attractedmorethan160regular,specialsession,anddemosession submissions from 21 countries/regions all over the world. All submitted papers have been reviewed through no less than workstation contributors or exterior reviewers, and so much of them have been reviewed by way of 3 reviewers. The assessment technique used to be very selective. From the full of 133 submissions to the most tune, forty three (32. 3%) have been permitted as typical papers, 22 (16. 5%) as brief papers. In all, 15 papers have been acquired for 3 detailed periods, that is through invitation in basic terms, and 14 submissions have been acquired for a demo consultation, with nine being chosen. Authors of permitted papers come from sixteen countries/regions. This quantity of the court cases comprises the abstracts of 3 invited talks and the entire common, brief, distinctive consultation and demo papers. The general papers have been classified into 9 sections: 3D mod- ing;advancedvideocodingandadaptation;face,gestureandapplications;image processing;imageretrieval;learningsemanticconcepts;mediaanalysisandm- eling; semantic video recommendations; and monitoring and movement research. 3 specific classes have been video research and occasion reputation, cross-X multimedia mining in huge scale, and cellular computing and functions. The technical programfeatured 3 invited talks, paralleloral presentation of the entire authorized commonplace and certain consultation papers, and poster classes for brief and demo papers.

Show description

Read or Download Advances in Multimedia Modeling: 16th International Multimedia Modeling Conference, MMM 2010, Chongqing, China, January 6-8, 2010. Proceedings PDF

Similar international books

Computer Safety, Reliability, and Security: 28th International Conference, SAFECOMP 2009, Hamburg, Germany, September 15-18, 2009. Proceedings

Computer-based structures became omnipresent commodities inside of our - vironment. whereas for a wide number of those structures equivalent to transportation platforms, nuclear or chemical vegetation, or clinical platforms their relation to security is clear, we frequently don't re? ect that others are as at once on the topic of dangers touching on damage performed to individuals or topic as, for instance, elevator keep an eye on or cellphones.

Extra resources for Advances in Multimedia Modeling: 16th International Multimedia Modeling Conference, MMM 2010, Chongqing, China, January 6-8, 2010. Proceedings

Example text

A method with temporal filtering was proposed to remove phantom volumes by a threshold on size, however, the small real objects would be wrongly removed by the lower threshold value, furthermore, this method does not work well when it is in the self-occlusion situation [9]. In most of the free viewpoints video synthesize based on visual hull method, consistent color or texture information were used in the removal of phantom volumes by color interpretation, however, it is useless for the scene that contains repetitive texture or similar color across a region [10].

G. the first value v2 , each time alternately. If such the NCC computed by views v2 and v3 is above some threshold W , that means the voxel is invisible for view v2 and the program is end (V E = Φ). Otherwise, we put v2 into the set V E(V E = {v3 , v2 }). Next, v4 is searched out and the mean NCC computed by views v2 , v3 , v4 is obtained again to determine whether v4 is put into the set V E or not. Until all the values in the set V B are searched, we acquire a final accurate visibility set V E. The algorithm excludes inexact views in V B through imposing photo-consistency constraint and produces the more accurate set of views.

We propose a direct quantitative analysis on the visibility methods by the data form. We demonstrate the performance of our approaches on synthetic dataset, and give a significant error analysis (Table 1). We define two types of errors: (1)Er1 which denotes the wrong rate of set V E (Section 2) relative to set V T (real visibility set). If there exists a camera X(X ∈ V E, X ∈ / V T ), the set V E is considered wrong, namely the visibility for the point is wrong. Er1 = MV E × 100%, NV E (5) where MV E denotes the number of the points which have the wrong visibility in V , and NV E denotes the number of all the points in V .

Download PDF sample

Rated 4.62 of 5 – based on 32 votes